A Perfect Skeleton, Zero Facts: Notes from a Night in the LCK Transfer Window

2:47 in the morning, November 2026, a rented apartment in Incheon. On screen sat the draft I had just finished building. It was beautiful in the purely technical sense: a tight headline, a clear source line, the one-sentence summary already written, the entity list fully populated, the timeline aligned to the day. Every cell of the skeleton had text in it. Except one. The cell marked "core information" was empty. I stared at that cell for eleven minutes. Two weeks earlier I had published an investigative piece on Chovy's move to Gen.G — three independent sources, a salary figure of 1.7 billion KRW per year, 1.2 million reads. That story began with a phone call close to midnight. It did not begin with a skeleton. That night was the reverse. I had a perfect skeleton and not a single fact to put inside it. I deleted the whole file. Four years later, looking back, I can finally name what happened: an empty information set, produced by a labelling machine that was very good at its job and an extraction stage that had failed. The transfer window is the season in which esports news gives itself permission to be loose. In the LCK, the window opens after the final and closes a few weeks later. Short window, large movement, and every passing day raises the value of publishing first. Everyone in the trade knows how stories are manufactured: an agent wanting to move a price, a player changing a profile picture, an anonymous account posting a status and deleting it, a fan page counting Instagram follows. All of that is data. None of it is information. What separates this window from earlier ones is volume. A 2026 article passed through an editor's hands. A 2026 tweet travels straight from an agent's phone to ten thousand people in four minutes, gets translated into four languages, and comes back to me labelled "according to Korean sources". The loop closes. On one occasion I received my own erroneous report back, cited as evidence. Then there is the second layer: aggregation systems. Most esports news readers consume today is not written by people. It is gathered by machines. A machine reads the headline, reads the tags, reads the section, and is very good at applying the label "esports". But when the body sits behind a JavaScript-rendered page, or inside an infographic image, or behind a paywall, the machine applies its label and returns an empty information set. Correct skeleton. No organs. That is where the danger starts. Kazan taught me: the ball and the Baron both refuse to forgive anyone who trembles. An empty set is a result, not an incident. When the input to a deep analysis contains not one verifiable claim, the correct response is not to infer an article that looks plausible. The correct response is to publish the null result, along with a diagnosis of where the chain broke. In my trade, that translates to: when there is no source, the report must say "no source yet". That sounds obvious. But look at the skeleton I received that night, and at any transfer draft sitting in your inbox. Nine cells need filling. The first is the game title. It sounds redundant, but it is a hard gate. League of Legends patches every two weeks; DOTA2 patches rarely and each patch is a seismic shift; CS2 lives on weapon and map update cadence; Valorant follows Riot's cycle but on an entirely different scale; Honor of Kings and Peace Elite run on seasons. A sentence like "team X got stronger after the patch" can be true in one title and meaningless in another. Without a game title, every comparison has the wrong unit. The second cell is one verifiable claim. This is the cell I have filled incorrectly more often than any other in my career. In 2026 I was sent to Russia to cover the World Cup. On 27 June, South Korea beat Germany 2-0 in Kazan. I wrote the piece in the state of a man who had just stolen a Baron. Five minutes later, Sweden beat Mexico 3-0 in the simultaneous fixture, and South Korea were eliminated on goal difference. That chain of events can be checked link by link: the score, the date, the venue, the goal difference. Because of that, the article still stands, even though the emotion inside it was the emotion of someone who had just been robbed of a pentakill. The Chovy piece was the same. Three sources, one salary figure, one analysis of how he would adapt to the new meta. The hard core lay in something another person could break open and test. An empty article has no core at all. And when an article has no hard core, every remaining sentence is decoration. I learned this from a real game: game two of the 2026 LCK Summer final, Khan on Jayce, finishing 7/0/4 with four towers destroyed. Without those numbers, my piece "The Jayce Epic" was just a string of adjectives. I found Jayce, and in him I found a whole generation of players thrashing for air. But Jayce had to be real first. The third cell is source identity and publication date. During a transfer window, "last week" and "three weeks ago" are two different stories: one is news, the other is history. A report with no date cannot be placed in any news cycle, and a source with no name caps every conclusion at the lowest confidence level — even when the content is correct. The fourth cell is region and league. The same team can sit at completely different levels across two titles. A League of Legends champion can be a wildcard in an FPS title. Cross-title generalisation is a methodological error, not a stylistic one. The fifth cell is quantitative data. In 2026 I sat and measured the Summer final VODs between Damwon Kia and DRX minute by minute. Canyon and Beryl controlled roughly 75 percent of map vision across the first fifteen minutes of every game. I did not take that figure from anyone; I counted it myself. Because of that, when I wrote that empty stadiums made them more focused, the sentence had something to stand on. Based on my experience following these matches, an analysis is only worth trusting when the writer states who measured, how they measured, and over how many games. Then comes the trap I consider the most dangerous in the whole skeleton: cells that depend on each other in a circle. In the analysis I received that night, the "entities involved" cell was instructed to derive itself "from the information points above" — and the points above were empty. The "source quality" cell was instructed to judge "from the source fields" — and the source fields were also empty. Two cells pointing at each other. In journalism this is a familiar enough error to have its own name: "according to sources close to the situation" leads to another article also "according to sources close to the situation", and the whole chain ends up resting on a single point — the original rumour. Three articles, one source, and nobody in three newsrooms knows whom they are quoting. The last trap, and the one readers see least: silence being read as safety. An empty risk table does not mean there is no risk. An article about a team that does not mention unpaid wages does not mean that team pays on time. During a transfer window, the three largest risks — unpaid wages, contracts that imprison players behind enormous release clauses, and wrist injuries — are all the kind that surface only when somebody agrees to speak. Nobody speaks, the table stays empty. And the empty table, in the eyes of a skimming reader, becomes a clean bill of health. The silence of Damwon was not emptiness, but the waiting room of history. The silence of a team owing wages is something else entirely, and the writer's job is to tell the two apart. This industry rewards the person who breaks a story first. Nobody rewards the person who kills one. A wrong article gets caught, corrected, remembered. An article with a correct skeleton gets shared, because it reads smoothly, it has a headline, it has sources, it has names — it lacks exactly one thing nobody checks: a fact. In five years of working in Incheon, I have seen more drafts killed for lacking a source than killed for being wrong. But the number of drafts published simply because "the skeleton already looked good" is larger than both. There is a kind of romanticising I once fell into and now have to remind myself about every week. In 2026 I called Summoner's Rift a field of poetry, each lane a stanza. Beautiful. But that beauty only stood because there was a keypress behind it. In the summer of 2026 I learned that legends do not need to finish first, they only need someone to retell them — and the reteller has to retell it right. Every Baron take has a corresponding long-range header, equal in obsession. An empty report is the same: it haunts the writer longer than a wrong one, because it is never corrected. Nobody corrects an article that never said anything. In the newsroom I proposed a three-line gate taped to the top of every transfer draft: which source, which date, which verifiable claim. If all three cannot be filled, the piece does not pass — even when the headline is written and the skeleton looks good. We call it the empty gate, and it saves me exactly one humiliation per season. For readers there is one question worth asking. If this transfer window passes without a single story about the team you follow, is that because the team is silent, or because none of us bothered to ask?